#!/usr/bin/env python

"""Nearest Neighbor Index module.
"""

__author__ = "Théo de la Hogue"
__credits__ = []
__copyright__ = "Copyright 2023, Ecole Nationale de l'Aviation Civile (ENAC)"
__license__ = "BSD"

from typing import TypeVar, Tuple, Any
from dataclasses import dataclass, field

from argaze import GazeFeatures

import numpy
from scipy.spatial.distance import cdist

@dataclass
class ScanPathAnalyzer(GazeFeatures.ScanPathAnalyzer):
    """Implementation of Nearest Neighbor Index algorithm as described in:

        **Di Nocera F., Terenzi M., Camilli M. (2006).**  
        *Another look at scanpath: distance to nearest neighbour as a measure of mental workload.*  
        Developments in Human Factors in Transportation, Design, and Evaluation.  
        [https://www.researchgate.net](https://www.researchgate.net/publication/239470608_Another_look_at_scanpath_distance_to_nearest_neighbour_as_a_measure_of_mental_workload)
    """

    size: tuple[float, float]
    """Frame dimension."""

    def __post_init__(self):

        super().__init__()

        self.__nearest_neighbor_index = 0

    def analyze(self, scan_path: GazeFeatures.ScanPathType):
        """Analyze scan path."""

        assert(len(scan_path) > 1)

        # Gather fixations focus points
        fixations_focus = []
        for step in scan_path:
            fixations_focus.append(step.first_fixation.focus)

        # Compute inter fixation distances
        distances = cdist(fixations_focus, fixations_focus)

        # Find minimal distances between each fixations
        minimums = numpy.apply_along_axis(lambda row: numpy.min(row[numpy.nonzero(row)]), 1, distances)

        # Average of minimun distances
        dNN = numpy.sum(minimums / len(fixations_focus))

        # Mean random distance
        dran = 0.5 * numpy.sqrt(self.size[0] * self.size[1] / len(fixations_focus))

        self.__nearest_neighbor_index = dNN / dran

    @property
    def nearest_neighbor_index(self) -> float:
        """Nearest Neighbor Index."""
        
        return self.__nearest_neighbor_index